OK, I think I see what's correct here, though I listen. "Where true," (where there are active light_rail cars being tested) delete the construction=light_rail tag and change the rail key from a value of construction to a value of light_rail.

So:
construction=(empty)
railway=light_rail

where true in the real world.

Sketching in stations, stops, etc...in the future as they evolve. This one is getting built in real-time, that's OK.

To wrap it up, for any given segment of rail in OSM where traffic is seen on the line, remove its construction tags (no longer railway=construction + construction=light_rail, rather simply railway=light_rail.

This is NOT the same as OSM denoting this is open for passenger service, it simply means "no longer under construction." When vehicles are being tested upon it, that's no longer construction (and Nat is correct that for safety, it'd be "more correct" to denote it as "active" rail). Not quite yet passenger rail, but no longer under construction either.

As you know an opening_date, add that tag with a good value. As you further build the "simply bare naked rail" line of infrastructure into something closer to a passenger line, include stations, stops and so on. That's what makes it a passenger route. You might 'ease in" to that by going to a PTv1 flavor light_rail, or go all the way up to PTv2 all at once. I'd do it in stages (v1, then v2), but that's just me. It IS helpful to learn how to build good v2 routes, even if you go "through" v1 to do so.

If I were to look for a parallel, I'd say this Bellevue-Redmond-area line is somewhere about 6 to 9 months ahead of where something like the Redlands line is going (somewhere in 2022, maybe 2023): "emerging local passenger rail" of a light-rail nature. Our wiki and map more-or-less capture this, but it's all rough. What is built in OSM can be used from OSM.

All that is being "said so" here is whether there is rail. Nothing is said of passenger service or open-to-public as something-else.

Testing of infrastructure happens. If there are test vehicles upon the rail, remove any construction tag. A route relation for passengers is in the future as of now. Theis is "alpha test of welds for structural integrity, safety and signals/switches" stuff, done with an un-humaned robot car with cameras going slow, I'd guess. So take off any proposed tag on that (those) element(s) of railway=rail.

This seems like it is whether a single railway element has a construction tag or not. Somebody said, "there is this kind of railway here." One segment at a time, it is or isn't construction=*. Say so. Leave on the tag, take off the tag. Up to you locally. One tag at a time, if you have to. Vehicles on it? In an "early test," yeah, new track has to have vehicles upon it for the first time someday.

To be clear, I do not claim to be authoritative in OSM. I merely do my best, perhaps "being a good example" as a high goal I do strive to achieve. I do so because I see other mappers strive to do their best, not be authoritative and "set a good example by being a good example." It works.

Thanks to all participants here, and I still have hope that some consensus can be achieved on these topics.

Quite a number of false allegations here by Adamant1 (again), is what I'll say about that. Though, maybe if I skate into "being generous territory" I give him the benefit of the doubt and say he has a misunderstanding or is offers us confusing language. I honestly don't understand much of his approach to mapping anyway, despite what might look like years of concerted efforts at large numbers of edits. Good luck to you, Adamant1, I wish you no ill will.

I thank Minh for pouring oil on these turbulent waters, hopefully calming them. (That's a metaphor for de-escalation, frequently necessary with one particular participant here). OSM is indeed a meritocracy (of mapping skills, sometimes of writing / communication skills); OSM is not "who you know."

<though it’s frustrating that “road” and “highway” are such overloaded terms.>

I agree that this is a crux of difficulty here. Keep up the excellent dialog, everyone, this is good stuff!

That's a fair observation!

Consensus can be difficult to achieve, but it isn't impossible. Again, I am cheered by seeing as much progress as I do.

Certainly, discussion is and will take place in Slack. I choose not to participate via that medium and am appreciative to learn of "California edition" momentum trickling to / being "cross-posted" (from Slack? OK) to talk-us. Maybe a talk-california mail-list could be born, but traffic (still) seems too low to fully justify that (now).

I applaud the momentum, discussion and serious movements towards good consensus (regarding trunk, et al re highway classification). This is difficult and good work, and exactly as Minh says, there remain some rough edges. That's to be expected and doesn't seem terribly contentious (nor does it need to be). Rome wasn't built in a day.

Hi Bradley and others: From his california/Draft HCG (linked by the US 2021 HCG) wiki, I understand Bradley to mean groupings of California as he describes them there (admittedly "arbitrarily," to use his word). Though I and my mapping do seem to fall "largely" into Bradley's "central" region of California, having deep roots in southern California and having travelled a great deal of Humboldt, Mendocino, Shasta, Trinity, Modoc and nearby "more farther northern" counties of northern California, I am quite familiar with Bradley's "northern" region as well. So, in short, I am very much a California-wide mapper, though I do focus on "my" various areas. These are broad strokes and our language use here can be difficult as we do that. No offense meant to anybody as they characterize somebody's statewide- (or "part-statewide'") mapping, and none taken.

However, I do defer to Adamant1's apparently deep familiarity with Shasta County, Redding and environs and leave mapping those areas to him and others, even as I might occasionally map in those areas (emphasis on occasionally) myself.

WRT the discussion-at-hand regarding "trunk," I am very much in "watching mode." Thank you to everybody for positive contributions to wiki, changeset comments and the map itself.

Thanks. We're getting there!

Well, OK. In my zeal to correct the map, I'll say my deletion was correct and your admonishment to do more research and find where it actually belonged is "more correct." Thank you for doing the leg-work on this.

Nice work!

Ah, thanks for the clarification.

If you say so. It's not too far from my house and I've driven it thousands of times, if there are new signs that "feather the edges" as you tag here, OK.

OK, I think I've got this (mostly?) correct: the segments of Blue which are shared by SDIY are now tagged railway=rail + usage=branch + railway:traffic_mode=mixed. Also, the Blue route=railway relation has a note tag of "SDIY is an operator only on segments tagged rail:traffic_mode=mixed."

ORM being VERY slow to render right now hinders my progress here; I'll take a look this weekend.

BTW, it's "trolley only" between Old Town and UCSD on the new Blue Line Extension, so that part for sure is properly tagged railway=rail. Are there specific segments (of Blue or Orange) you think should also be so tagged?

(I'm sure none of Green should be, it's "all trolley" too).

Taking a look over the next day or two. I believe you are correct: segments of Blue (and parts of Orange not currently so tagged?) between the Santa Fe Station and the new Mid-Coast addition (to UCSD, which are purely light_rail) likely ARE railway=rail, usage=branch. Thing is, I'm not sure where the split to "light_rail only" is. I think it is near the Green branch north of Old Town (station), but I'm not sure where.